Agent Skills: graphrag-evaluation

Use when evaluating GraphRAG system quality across knowledge graph completeness, retrieval relevance, answer correctness, and reasoning verification. Invoke when user mentions evaluate GraphRAG, quality metrics, benchmark, hallucination reduction, answer correctness, multi-step reasoning evaluation, test my GraphRAG, or measure RAG performance. Provides evaluation frameworks, metric selection, and testing protocols.